This application for a K01 award is to facilitate Dr. Arpita Chattopadhyay's transition to an independent investigator in the field of epidemiology of aging and long-term care with a focus on Medicaid policy. Her long-term goal is to develop an interdisciplinary program linking biomedical research on aging and disability with behavioral research in economics and sociology that focuses on utilization and financing of long-term care among the elderly. The training goals are (1) to acquire substantive knowledge in sociology and economics of aging and long-term care policies; (2) to acquire new skills in developing stochastic models of disability; and (3) to develop a frame work for estimating the determinants of disability, long-term care use, and financing. The mentoring team includes Dr. Andrew Bindman, an experienced health services researcher; Drs. Robert Newcomer and Charlene Harrington, experts in long-term care issues; Dr. Peter Bacchetti, expert in stochastic processes in chronic disease modeling; and Dr. Ronal Lee, who specializes in the demography of aging The impact of changing racial and ethnic composition on the demand for long-term care and utilization of Medicaid for financing long-term care has not been studied. This research will overcome this shortcoming by incorporating race and ethnicity into projection models for long-term care demand and Medicaid use among the elderly. The research will address two of the priority areas at the NIA, namely aging and health /healthcare disparities and efficiency of the health system. Drawing upon data from sample surveys and administrative sources on service use, this research proposes four specific aims that correspond to four structural determinants of future long-term care use and financing. Aim 1 will determine racial and ethnic differences in age-specific prevalence of disability. Aim 2 will determine racial and ethnic differences in age specific use of long-term care for a given level of disability. Aim 3 will determine the racial and ethnic differences in use of Medicaid among the disabled elderly. Aims 1-3 will then be combined and used to obtain stochastic projection of elderly Medicaid beneficiaries in long-term care (Aim 4). A detailed microsimulation model incorporating behavioral and structural determinants of using Medicaid financed longterm care will be proposed for an R01 project at the end of the K01.